Brain age gap estimation using attention-based ResNet method for Alzheimer's disease detection.

Area of Science:

• Neuroimaging
• Artificial Intelligence in Medicine
• Neurology

Background:

• Neurodegenerative diseases like Alzheimer's pose significant public health challenges.
• Accurate biomarkers are crucial for early diagnosis and effective disease management.
• Brain age estimation from MRI offers a potential avenue for identifying neurological decline.

Purpose of the Study:

• To develop and validate a novel deep learning model for accurate brain age estimation.
• To compute the brain age gap (BAG) as a potential biomarker for Alzheimer's disease (AD).
• To assess the model's generalizability and performance in distinguishing between cognitively normal and AD individuals.

Main Methods:

• A novel attention-based ResNet method, 3D-Attention-Resent-SVR, was developed for brain age estimation.
• The model was trained and tested on a combined dataset of 3844 individuals from four public sources.
• Brain age gap (BAG) was computed to differentiate between Cognitively Normal (CN) and Alzheimer's disease (AD) groups.

Main Results:

• The model achieved a mean absolute error (MAE) of 2.05 for brain age gap estimation on the combined dataset.
• Excellent generalizability was demonstrated with an MAE of 2.4 when trained on three datasets and tested on a separate one.
• Using BAG as a sole biomarker, the model achieved 92% accuracy and an AUC of 0.87 in AD detection on the ADNI dataset.

Conclusions:

• The proposed 3D-Attention-Resent-SVR model accurately estimates brain age and the brain age gap.
• The brain age gap (BAG) serves as a potent biomarker for early Alzheimer's disease detection.
• This approach holds significant potential for early diagnosis and monitoring of neurodegenerative diseases.